Sudden cardiac death (SCD) remains a serious outcome in hypertrophic cardiomyopathy (HCM), with an estimated incidence of 1%/year. Its risk prediction stays challenging for clinicians. Implantation of cardioverter-defibrillator is recommended with a strong level of evidence when the HCM risk score is over 6%. However, the ESC HCM risk score is criticized for its moderate diagnostic performance (C-index 0,70), and its low sensibility. Recent studies suggest that low PALS (peak atrial longitudinal strain) is associated with poor prognosis in HCM. Artificial intelligence, wich is experiencing an unprecedented boom, may be an interesting tool in association to left atrial strain in Prognostication. This preliminar study aimed to evaluate the value of the left atrial strain coupled with artificial inteligence algorythm in the prediction of suddent cardiac death in hypertrophic cardiomyopathy. This restrospective multi-center cohort (Marseille la Timone, Bordeaux and Rennes) included 431 patients with sarcomeric HCM from 2007 to 2018. The average follow-up time was 57 months. The primary endpoint was all cause death, ventricular arythmias, recovered SCD and internal electric shock. Analyses were entrusted to the LIS laboratory specialized in Computer Science at Aix-Marseille University. Sixty-six percent of men composed the cohort with a mean age of 52 ± 16 yo. Sarcomeric mutations were found in 52% of them, the mean left ventricular ejection fraction was 63 ± 16, 23% had significant ventricular obstruction. The mean PALS was 25 ± 12%. In the multivariate analysis, PALS 4 chamber, left ventricular global longitudinal strain and syncope were associated with the primary endpoint, (respectivly HR 2,48 IC95% [1,09–5,63]; P 0,030; HR 1,13 IC95% [1,03–1,25]; P 0,012; HR 0,95 IC95% [0,92–0,99]; P: 0.023). The random Forrest algorithm was able to successfully classify the patient (occurrence of primary endpoint or not) in 92,2% of the cases. Left atrial strain coupled with the Random Forrest algorithm could offer an interesting implementative value for the prediction of sudden cardiac death in hypertrophic cardiomyopathy. Further analysis are required to define its sensibility and specificity.